Re Repr presentational Di Dimen ensions Com omputer Science c - - PowerPoint PPT Presentation

CPSC 322, Lecture 2 Slide 1

Re Repr presentational Di Dimen ensions

Com

• mputer Science c

cpsc sc322, Lecture 2 2 (Te Text xtboo

• ok

k Chpt1)

Ma May, y, 1 16, 2 2017

CPSC 322, Lecture 2 Slide 2

Lectu ture re Ov Overv rvie iew

• Recap

ap fr from l las ast l lecture

• Representation and Reasoning
• An Overview of This Course
• Further Dimensions of Representational

Complexity

CPSC 322, Lecture 2 Slide 3

Co Cours rse Ess ssenti tial als

• Cou
• urse

se web-page ge : : CHECK IT OFTEN!

• Te

Text xtboo

• ok: Available online!
• We will cover at least Chapters: 1, 3, 4, 5, 6, 8, 9
• Con
• nnect: discussion board, grades
• AI

AIsp space : online tools for learning Artificial Intelligence http://aispace.org/

• Lecture slides…
• Midterm exa

xam, , planning t g to

• have in on
• n Wed J

Jun 7 7 (will have a doo

• odle on
• n piazz

zza)

CPSC 322, Lecture 2 Slide 4

Age gents ts ac acti ting g in in an an envi viro ronme ment

Representation & Reasoning

CPSC 322, Lecture 2 Slide 5

Lectu ture re Ov Overv rvie iew

• Recap from last lecture
• Repr

presentat atio ion n an and d Reas asonin ing

• An Overview of This Course
• Further Dimensions of Representational

Complexity

CPSC 322, Lecture 2 Slide 6

What t do

• we n

need to to re repre rese sent t ?

• Th

The environ

• nment /wor
• rld : What different configurations

(st states s / pos

• ssi

sible wor

• rlds) can the world be in, and how

do we denote them? Chessboard, Info about a patient, Robot Location

• Ho

How the w wor

• rld wor
• rks

ks (we will focus on)

• Con
• nst

straints: s: sum of current into a node = 0

• Causa

sal: what are the causes and the effects of brain disorders?

• Ac

Action

• ns preconditions and effects: when can I press

this button? What happens if I press it?

CPSC 322, Lecture 2 Slide 7

Cor

• rre

resp spon

• ndin

ing g Reaso sonin ing g Task sks s / P Pro roble lems

• Con
• nst

straint Satisf sfaction

• n – Fi

Find st state that sa satisf sfies s se set

• f
• f con
• nst

straints.

• s. E.g., What is a feasible schedule for

final exams?

• An

Answ swering g Query – Is Is a gi given prop

• pos
• sition
• n true/like

kely gi given what is s kn know

• wn? E.g., Does this patient suffers

from viral hepatitis?

• Planning

g – Fi Find se sequence of

• f action
• ns

s to

• reach a

a go goal st state / maxi ximize ze utility. . E.g., Navigate through and environment to reach a particular location. Collect gems and avoid monsters

CPSC 322, Lecture 2 Slide 8

Repre rese senta tati tion

• n an

and Reas ason

• nin

ing g Sy Syst stem

• A

(represe sentation

• n) langu

guage ge in which the environment and how it works can be described

• Computational (reaso

soning) proc

• cedures to compute a

solution to a problem in that environment (an answer , a sequence of actions) Bu But the choice of an appropriate R&R system depends on a key property of the environment and of the agent’s knowledge

CPSC 322, Lecture 2 Slide 9

Det eter ermi mini nist stic ic vs

• vs. Sto

Stoch chas asti tic c (U (Unc ncer erta tain in) Dom

• mai

ains

• Se

Sensi sing g Un Uncertainty: Can the agent fully observe the current state of the world?

• Effect Un

Uncertainty: Does the agent knows for sure what the effects of its actions are? Chess Poker Factory Floor Doctor Diagnosis Doctor T reatment

Cl Clic icke ker r Qu Quest stio ion: Ch Chess ss an and Pok

• ker
• A. Poker and Chess are both stochastic
• B. Chess is stochastic and Poker is deterministic
• C. Poker and Chess are both stochastic
• D. Chess is deterministic and Poker is stochastic

CPSC 322, Lecture 2 Slide 10

Sto tochas asti tic if at at l leas ast o t one of th these is tr true

• Se

Sensing Uncerta tainty ty: Can the agent fully observe the current state of the world?

• Effect U

t Uncerta tainty ty: Does the agent knows for sure what the effects of its actions are?

CPSC 322, Lecture 2 Slide 1 1

Dete term rmin inis isti tic vs

• vs. Sto

Stochas asti tic Dom

• mai

ains

Historically, AI has been divided into two camps: those who prefer representations based on log

• gic and those

who prefer prob

• bability.

A few years ago, CPSC 322 covered logic, while CPSC 422 422 introduced probability:

• now we introduce both representational families in

322, and 422 goes into more depth

• this should give you a better idea of what's included

in AI No Note: Some of the most exciting current research in AI is actually building bridges between these camps.

CPSC 322, Lecture 2 Slide 12

Lectu ture re Ov Overv rvie iew

• Recap from last lecture
• Representation and Reasoning
• An Ov

Overvi view of

• f T

This is Co Course

• Further Dimensions of Representational

Complexity

CPSC 322, Lecture 2 Slide 13

Mod

• dul

ules es we' e'll ll co cove ver r in in th this is co cour urse se: R&R &Rsy sys

Environ

• nment

Prob

• blem

Query Planning Deterministic Stochastic Constraint Satisfaction Search Arc Consistency Search Search Logics STRIPS Vars + Constraints Value Iteration

• Var. Elimination

Belief Nets Decision Nets Markov Processes

• Var. Elimination

Sta tati tic Se Sequenti tial al Representation Reasoning Technique

CPSC 322, Lecture 2 Slide 14

Lectu ture re Ov Overv rvie iew

• Recap from last lecture
• Representation
• An Overview of This Course
• Fu

Further Dim imensio ions of f Repr presentat atio iona nal l Co Compl plexit ity

CPSC 322, Lecture 2 Slide 15

Dimensions of f Repr presentat ationa nal Co Compl plexity

We've already disc scuss ssed:

• Problems /Reasoning tasks (Static vs. Sequential )
• Deterministic versus stochastic domains

Som

• me ot
• ther impor
• rtant dimensi

sion

• ns

s of

• f com
• mplexi

xity:

• Explicit state or propositions or relations
• Flat or hierarchical
• Knowledge given versus knowledge learned from

experience

• Goals versus complex preferences
• Single-agent vs. multi-agent

CPSC 322, Lecture 2 Slide 16

Exp xpli licit it Sta State te or

• r pro

ropos

• sit

itio ions

How do we model the environment?

• Y
• u can enumerate the st

states of the world.

• A

state can be described in terms of features

• Often it is more natural to describe states in terms of

assignments of values to features (variables).

• 30 binary features (also called propositions) can represent

230= 1,073,741,824 states.

Mars s Exp xplor

• rer Exa

xample Weather Temperature LocX LocY

CPSC 322, Lecture 2 Slide 17

Exp xplicit t St State te or

• r pro

ropos

• siti

tion

• ns

s or

• r re

relati tion

• ns
• States can be described in terms of ob
• bjects and

relation

• nsh

ships.

• There is a proposition for each relationship on each

“possible” tuple of individuals. Un Universi sity Exa xample Registred(S,C) Students (S) = { } Courses (C) = { }

Cl Clic icke ker r Qu Quest stio ion

One binary relation (e.g., likes) and 9 individuals (people). How many states?

• A. 812
• B. 102
• C. 281
• D. 109

CPSC 322, Lecture 2 Slide 18

I changed same-nationality to likes because if you reason on the meaning of same-nationality the states are less, they are 236

Co Comp mple lete te Exa xamp mple le

CPSC 322, Lecture 2 Slide 19

CPSC 322, Lecture 2 Slide 20

Fl Flat at or

• r hie

iera rarc rchic ical al

Is it useful to model the whole world at the same level of abstraction?

• Y
• u can model the world at one level of abstraction: flat
• Y
• u can model the world at multiple levels of abstraction:

hierarchical

• Example: Planning a trip from here to a resort in Cancun, Mexico

CPSC 322, Lecture 2 Slide 21

Knowle ledg dge giv iven vs vs. . knowle ledg dge le lear arned fr d from expe perie ience

The agent is provided with a model of the world once and far all

• The agent can learn how the world works based on

experience

• in this case, the agent often still does start out with some

prior knowledge

CPSC 322, Lecture 2 Slide 22

Goa

• als

ls ve vers rsus s (c (com

• mple

lex) x) pre refe fere rences

An agent may have preferences

• e.g., there is some preference/uti

tility ty functi tion that describes how happy the agent is in each state of the world; the agent's task is to reach a state which makes it as happy as possible

An agent may have a go goal that it wants to achieve

• e.g., there is some sta

tate te o

• r set o

t of sta tate tes of the world that the agent wants to be in

• e.g., there is some propositi

tion

• n or set o

t of propositi tion

• ns that the

agent wants to make true What beverage to order?

• The sooner I get one the better
• Cappuccino better than Espresso

Preferences can be complex…

CPSC 322, Lecture 2 Slide 23

Si Singl gle-ag agent t vs

• vs. Mult

ltia iage gent t dom

• mai

ains

Does the environment include other agents? Everything we've said so far presumes that there is only

• ne agent in the environment.
• If there are other agents whose actions affect us, it can

be useful to exp xplicitly mod

• del their go

goals s and b beliefs rather than considering them to be part of the environment

• Other

Agents can be: coo

• operative, com
• mpetitive, or a bit of
• f

bot

• th

CPSC 322, Lecture 2 Slide 24

Dimensions of f Repr presentationa nal Compl plexity in in C CPSC3 C322

• Reasoning tasks (Constraint Satisfaction /

Logic&Probabilistic Inference / Planning)

• Deterministic versus stochastic domains

So Some ot

• ther impor
• rtant dimensi

sion

• ns

s of

• f com
• mplexi

xity:

• Explicit state or features or relations
• Flat or hierarchical
• Knowledge given versus knowledge learned from

experience

• Goals vs. (complex) preferences
• Single-agent vs. multi-agent

CPSC 322, Lecture 2 Slide 25

In cla lass ss ac acti tivi vity ty

• Work in pair searching the web to find a coo
• ol exa

xample

• f
• f fielded (or
• r exp

xperimental AI AI age gents) s) you found

• Try to find something different from the usual

suspects (IBM Watson, Apple’s Siri and Microsoft’s Cortana)

Tr Try t to a answer t the fo following qu questions (t (take n notes)

• 1. What does the application do? (e.g,. control a spacecraft,

perform medical diagnoses, provide intelligent help for computer users, shop on eBay)

• 2. List some of the application : goals /preferences; observations

that it needs about the environment; types of actions that it performs

• 3. What AI te

technologies does the application use (e.g,. belief networks, Markov models, semantic networks, heuristic search, constraint satisfaction, planning)

• 4. Why is it intelligent? Which aspects make it an intelligent

system?

• 5. Is it an experimental system or a fielded system (i.e., used in a

real world setting)?

• 6. Is evidence provided on how well does the application perform?

CPSC 322, Lecture 2 Slide 26

CPSC 322, Lecture 2 Slide 27

Ne Next xt par art

• Assignment 0 due: submit electronically and you
• u can't

use se late days

• Hi

Hint: : AA AAAI AI is s the main AI AI ass ssociation

• Come to class ready to discuss the two
• exa

xamples s of

• f

fielded AI AI age gents s you found

• I'll show some pictures of cool applications in that

class

• Read carefully Section 1.6 on textbook: “Example

Applications”

• The Tutoring System
• The trading agent
• The autonomous delivery robot
• The diagnostic assistant

CPSC 322, Lecture 2 Slide 28